140

I came across this on the web. It's supposed to be funny, but we all know better :)

enter image description here

So, why are the numpad layouts different, and what are the reasons behind each?

12
  • 5
    +1 to you sir; I too have so often wondered this and been frustrated by it. Commented Jan 28, 2012 at 17:38
  • 28
    What's really going to bake your noodle later on is, why is the zero placed at the bottom in both layouts?!
    – Rahul
    Commented Jan 29, 2012 at 11:52
  • 1
    @Rahul I'm actually pretty sure about that one :). On the kbd it's because it's the most frequently used digit (in "real-life" numbers but not in phone numbers), and on the phone - so as to not break apart the row of *0# - which would be bad for all kinds of reasons. And you're not going to place that row above the other digits, because it's hardly ever used. Commented Jan 29, 2012 at 17:16
  • 2
    There is also now a nice numberphile video which summarizes the history behind this "inconsistency". My personal theory is that numbers used for calculations, or just as a symbol collection are different enough use cases, that they might warrant different interfaces. Commented Apr 17, 2015 at 22:36
  • 3
    @JoeMcGrath I'm only asking because it sounds like the question of someone who's never used a stationary phone, but grew around mobile or at least wireless phones. The fingers were heavily used on stationary phones (and actually the thumb - not so much). Commented Aug 21, 2015 at 15:31

7 Answers 7

88

There's this humongous article called Keyboard Trivia that has collected many of the theories and stories. The summary of facts:

  • Touch-tone key pad was designed to mimic the rotary dial with the "1" on top and the 7-8-9 on the bottom, and AT&T conducted user testing to confirm that this configuration helped eliminate dialing errors.
  • By the time when the touch-tone telephone was being designed in the late 1950s, the calculator and adding-machine designers had already established a layout that had 7, 8 and 9 across the top row.
  • Back then, the industry-standard typical calculator had nine columns of numbers, with 10 numbers in a column, the lowest digits at the bottom, starting with 0 and moving up to 9, and was basically a mechanical adding machine that closely resembled a cash register.
  • It is common practice today to use the telephone-keypad layout when designing new products that utilize a keypad, such as Automated Teller Machines.
  • When Bell Labs began exploring keypad layouts in the late 1950s they contacted all of the leading calculator manufacturers to find out why they had chosen to put low numbers at the bottom and high numbers at the top rather than the other way around. The answer, apparently, was a big shrug. It turns out that decision was largely arbitrary: no one had done any research about which layout was most convenient for users. Still, when it came time to place a numeric keypad on a computer keyboard, the calculator model with 7-8-9 at the top prevailed.

There's also a theory that phone engineers wanted to slow down people who were fast at entering numerical data, which would jam lines and produce dialing errors, so they reversed the layout. However, records of AT&T Labs' research clearly invalidate it.

12
  • 1
    In a left to right / top to bottom culture the telephone layout is more logical...123,456,789
    – PhillipW
    Commented Jan 28, 2012 at 20:26
  • 15
    PhilipW: tell that to an accountant that is subject to Benford's law en.wikipedia.org/wiki/Benford's_law Commented Feb 1, 2012 at 6:25
  • 2
    isn't it a hypothesis rather than a theory about the phone engineers, i.e. something that might be true, or makes sense, sounds possible etc, but has no evidence (yet), whereas theories are often the best interpretation of a wealth of evidence and actually very close to fact in many cases ?
    – Toni Leigh
    Commented Jan 16, 2014 at 8:26
  • 2
    @PhillipW: Whether 1 to 9 goes up or down, 0 should be closer to 1 than to 9 to be more logical, as 0…9 is a natural and logical increasing order, while 1…0 is not.
    – musiphil
    Commented Apr 6, 2016 at 21:11
  • 2
    @musiphil: best argument ever heard
    – Bernhard
    Commented Dec 14, 2016 at 10:33
15

I have a reasonable explanation. Since I come from mathematics background, this is the first thing that comes to mind.

Calculators

You probably have heard of the Benford's law, also known as the "First digit law". It says, that digits 1,2,3 occur more often than 7,8,9 when dealing with random, arbitrarily big numbers, which numbers that users type into calculators, in fact, are.

The graph is something like this and shows heavy bias against the digit "1". Therefore it's sensible to put digits 1, 2, and 3 to the bottom row, as it's closer to the enter and also user reaches them easier.

enter image description here Benford's law (wikipedia)

As pointed out in the comments, 0 tends to occur very frequently as well, due to rounding using the decimal number system. This is why the zero key is especially large.

Phone numbers

Since phone numbers have a fixed number of digits which randomly scattered, the Benford's law doesn't apply here, which means all the digits occur with similar frequencies. This might be the reason that the more natural ordering is used, since there's no reason to complicate.

3
  • I am sorry but could not understand the connection between the question and your answer. What about the zero btw?
    – Abektes
    Commented May 27, 2014 at 14:38
  • @Abektes zero is absurdly more commonly used than all the other digits, at least when used on a calculator, because many numbers have trailing zeroes.
    – Hakanai
    Commented Nov 13, 2022 at 7:45
  • @Hakanai: Good point, I updated the answer.
    – Rok Kralj
    Commented Nov 14, 2022 at 22:54
13

One additional possible explanation is that with the calculator layout many calculations can be performed along the bottom two rows and the plus button only; zeroes, decimal points and the enter/equals button.

Since the main buttons are all at the bottom, it makes reasonable sense that the other numbers would count up from the bottom too.

I suspect (but have no proof) that the phone keypad layout would make typing on a calculator more difficult for most simple calculations.

3
  • 1
    The proof lies in Fitt's Law plus Zipf's Law. Zipf's Law states that low numbers (0,1,2,3) are much more common in random numeric data than high numbers. Fitt's Law states that it takes longer to hit buttons that are further away...and hand position means the bottom row is closer. Commented Jan 14, 2013 at 20:28
  • 2
    @AlexFeinman: do you mean Benford's law? Zipf's law is about the distribution of words, not numbers.
    – Kit Grose
    Commented Jan 17, 2013 at 0:00
  • 2
    I do! Thanks for the correction. The two laws are functionally similar, but have different implications. Commented Jan 17, 2013 at 13:54
4

In the early sixties I took a year out of university and got a job in the internal audit department at the head office of a large insurance company in Toronto.

A substantial part of my job involved using an electromechanical calculator to check figures submitted by various departments. The keyboard layout was identical to present computer layouts as well present digital calculators. Index finger was for the left column, next two for adjacent ones. Thumb did zeros exclusively. I became very fast at this touch based skill, as my eyes were on the numbers on paper (there was no visual feedback - just the numbers that were printing on the tape, but you didn't look at that. Mistakes were extremely rare, and required a redo of the entire document. This was also in the days of rotary phones.

Imagine my shock when reversed digital keys appeared on telephones. My skill was not transferable to them. The decision by Bell Labs, or whomever, to reverse the layout was thoughtless at best, stupid at worst.

3

The phone design of 123 Top, and 789 Bottom was merely based on the research done by the Bell's labs:

"This was due to research conducted by Bell Labs using test subjects unfamiliar with keypads. Comparing various layouts including two-row, two-column, and circular configurations, the study concluded that while there was little difference in speed or accuracy between any of the layouts, the now familiar arrangement with 1 at the top was the most favourably rated."

So with no (or not enough) factual studies to rationale the designs, we can say that:

Phones:

Push-button or Touch-tone keypad designs are merely based on the Bell's labs research, and I would add it is a logical order when no studies proving the opposite existed at that time. Reading is Top to Bottom, Left to Right for those who designed it (in USA), so it starts 1,2,3 Top...to 7,8,9 Bottom [Total of 9 digits] we have three rows with three digits in each.

Zero [a unique digit]:

Zero is the digit number 10 (does not fit in a 3x3 matrix), it has to be located somewhere. Create another row (of 3 digits) below as you going Top to Bottom. Now if you put it on the left or right corner, you need a reason to do so. Locating it in the middle makes it a unique location (it is a unique digit) and easy to spot [my personal point of view].

Another point: if you take a look at the Rotary-dial phones, notice that they mostly have the Zero as the first digit located bottom middle, so Push-button design kept the zero in the same location as the Rotary-dial design, maybe?

*Some Rotary-dial phones has the zero a little to the right.

enter image description here

*They added the [ *,#] later on (added more functions and fill out the 4th row).

Mobile phones just matched early phones' design.

Keyboards:

As answered above regarding Benford's law:

"Briefly explained, Benford's Law maintains that the numeral 1 will be the leading digit in a genuine data set of numbers 30.1% of the time; the numeral 2 will be the leading digit 17.6% of the time; and each subsequent numeral, 3 through 9, will be the leading digit with decreasing frequency."

Numpad on keyboards were not designed with the goal to make phone calls, they were rather designed for accounting and calculations purposes (helpful in applications were numbers are heavily used). Adopting the Benford's law, it makes more sense to locate the most used numbers close to the human fingers on a keyboard (1,2,3 Bottom...7,8,9 Top).

Notice that Zero is always placed bottom no matter the numbers layout. Users just got used to it there? maybe.

Personally, I saw many bank employees or generally accountants using the numpad on keyboards very fast without the need to look at the keyboard, maybe they are just used to it. What if the design is Top - Bottom will they be as fast? Well, not according to Benford's law.

EDIT: Just adding this info from Wikipedia Rotary dial

Different pulse systems are used, varying from country to country. For example, Sweden uses one pulse to signal the number zero, and 10 pulses to signal the number nine. New Zealand uses ten pulses minus the number desired; so dialing 7 produces three pulses. In Norway, the North American system with the number '1' corresponding to one pulse was used, except for the capital, Oslo, which used the same "inverse" system as in New Zealand.

For this reason, the numbers on the dial are shifted in different countries, or even in different areas of one country, to work with their system because of the difference of the number arrangement on the dial. The dial numbering can occur in four different formats, with 0 adjacent either to the 1 or the 9, and the numbers running in ascending or descending order, with either the 0, 1 or 9 being closest to the finger stop.

4
  • 1
    Personally, I always thought the 0 at the bottom was related to the original way of using phones, were in many cases you had to dial 0 in order to get a communication, while first phone numbers were chronological so Benford's law applied. Then it evolved from there with the * and # added at the sides of 0 number. As for calculators, it just seems logical: 0, dot, 1,2,3.... . Just a theory on a possible explanation, not that it holds a lot of merit or that I have any documentation to support it. Your answer kind of supports this theory, at least in part
    – Devin
    Commented Mar 19, 2018 at 19:08
  • Good point. So Zero has been always a unique number with unique functionality. Phone systems especially inside organizations till now use Zero for special functions. Push Zero for the operator, Zero to speak to representative...etc. I just wonder why the Rotary-dial although had Zero at the bottom, it is not followed by 1,2,3...but 9,8,7.... Or If we look at it differently they started with 1,2,3... top (from the Right not left tho) and ended with Zero.
    – Moath
    Commented Mar 20, 2018 at 13:40
  • @Mo'ath - actually, if you understand the mechanics of the early direct-dial phone system (pre-touch-tone, and rotary dials rather than buttons), 0 at the bottom makes sense - on rotary-dial phones, 0 was after 9 because it generated ten pulses on the line for dialling - so it's not really a zero, it's a ten, even in these touch-tone days. Commented Mar 20, 2019 at 18:22
  • I added it to my answer above. A quick search produced this: "Different pulse systems are used, varying from country to country. For example, Sweden uses one pulse to signal the number zero, and 10 pulses to signal the number nine. New Zealand uses ten pulses minus the number desired; so dialing 7 produces three pulses. In Norway, the North American system with the number '1' corresponding to one pulse was used, except for the capital, Oslo, which used the same "inverse" system as in New Zealand"
    – Moath
    Commented Mar 20, 2019 at 19:29
0

This phone was the standard model in Norway during the 1980s. Note the calculator layout.

Norwegian phone

Norwegian coin operated phone, year unknown, also with calculator layout:

enter image description here

-1

Also: you tend to hold the device at a different angle. You might use all five fingers on a keyboard numpad, but on your phone only your thumb (or the index finger of the hand not holding the phone)..

1
  • Does this fundamentally affect the layout choice though? In both situations, I need to use all of the keys available, and experience no bias toward higher or lower numbers. Commented Jan 20, 2014 at 6:58

Not the answer you're looking for? Browse other questions tagged or ask your own question.